|type de publication
||article dans une revue internationale avec comité de lecture
|date de publication
||Danquigny Charles; Ackerer Philippe; Carlier Jean-Philippe|
||Journal of Hydrology (J Hydrol)
| || |
||105 – 115
||Tracer tests have been performed on a 3D tank of dimensions 560×100×100 cm3. These experiments at laboratory scale are used to define effective hydraulic conductivity and macrodispersivity. The tracer tests have been performed on two kinds of heterogeneous porous material: a channel structured medium, with channels crossing the whole tank, and a statistically correlated random structure. The statistically correlated field was first established by a sequential type generator with a prescribed exponential covariance. The theoretical random field is then modified according to the available sand hydraulic conductivities. The obtained LnK variance is 1.03 and the integral scale 23.1 cm.
Results obtained concerning the channel structured media show that it behaves like a stratified medium from an hydrodynamical and mass transfer point of view. The effective hydraulic conductivity is equal to the arithmetic mean and the variance of the concentration is proportional to t2, t being the mean travel time of the tracer displacement..
For the correlated random field, the effective hydraulic conductivity falls between by the geometric and the arithmetic mean. No match with theoretical values have been found because the boundary conditions effects cannot be neglected due to the dimension of the tank compared to the integral scale. Similar conclusions have been obtained for macrodispersivity. The tracer tests could be simulated by a 1D advection–dispersion equation but with a significant higher macrodispersivity than predicted by stochastic theories for infinite media.
||Tracer test; Groundwater; Effective parameters; Hydraulic conductivity; Dispersivity